o cin is cin2tab's native format, it's from xcin originally, but now CCE's
   cin file has lots of different keywords.

/**************************************************************************************************
 *
 *          CIN  (Chinese Input File) Format
 *
 *     # This is a comment line 
 *         # # is used for a comment line, cin2tab will skip this line
 *         #
 *     [%gen_inp]
 *         # This is for compatibility with XCIN, not used 
 *         #
 *     [%encoding|%coding N]
 *         # specifiy the encoding of this input table, otherwise you need to use command
 *         # line options -gb|-gbk|-big5|-jis|-ksc etc to specify it
 *         #
 *     [%ename ENGNAME]
 *         # specify the English name for this input method
 *         #
 *     [%cname|%prompt CHNAME]
 *         # specify the Chinese name for this input method
 *         #
 *     [%selkey xxxxxxxx]
 *         # specify the select keys(used to select the candidate), usually 1234567890(default)
 *         # needs to be printable, can't accept \0 \x format
 *         #
 *     [%dupsel n]
 *         # specify the number of select keys, or cin2tab will get it from &selkey
 *         #
 *     [%endkey yyyy]
 *         # This is for compatibility with XCIN, not used
 *         #
 *     [%pageprevkey xxxx]
 *         # specify the PrevPage keys, default is ",<-[" optional
 *         # needs not to be printable, can accept \0 \x format, max 7 keys
 *         #
 *     [%pagenextkey xxxx]
 *         # specify the NextPage keys, default is ".>=]" optional
 *         # needs not to be printable, can accept \0 \x format, max 7 keys
 *         # (Big5 tradition, SPACE is also a NextPage key for Zhuyin/CangJie/Simplex)
 *         #
 *     [%selfirstkey xxxx]
 *         # specify the key used to select first candidate, default is SPACE
 *         # needs not to be printable, can accept \0 \x format, max 3 keys
 *         # for example: SPACE  \032
 *         #
 *     [%wildcardkey x]
 *         # specify the WildCard key which can match any input keys
 *         # default is '?'   can't be one of the input key or select keys
 *         #
 *     [%maxpress n]
 *         # specify the maximum keys, maxpress will be generated by parsing the CIN file
 *         # the %maxpress is just for compatibility and information only
 *         #
 *     [%last_full|%auto_select 0|1]
 *         # if it's 1, when maxpress key is input and there is only one candidate, CCE will
 *         # automatically output this candidate, good for something like WuBi
 *         #
 *     [%key_prompt 0|1]
 *         # if it's 1, CCE will display the uninput key for the candidates, like in Windows
 *         # good for sth like WuBi/CangJie. Default is off
 *         #
 *     [%left_shift 0|1]
 *         # if it's 1(default), you can use Left-Shift to temporary enter English mode
 *         # press Left-Shift again to go back to Chinese input mode
 *         #
 *     [%prefix_match 0|1]
 *         # if it's 1(default), display candidates with the input keys as prefix(prefix match)
 *         # otherwise, it's exact match, only display candidates with the current input keys
 *         #
 *     [%key_len|%keylen n]  n=5/6, 6 is default
 *         # CCE will use either 5 or 6 bits to encode each key, so one 32 bits integer can 
 *         # hold 6 or 5 keys. If you omit this, cin2tab will calculate it from %keyname
 *         # CCE uses 1-2 integers to hold the keys(so max 10-12 keys). It will determine
 *         # that by %maxpress and %keylen.
 *         #
 *         
 *     %keyname begin
 *       <INPUT_KEY1> <DISPLAY1>
 *       <INPUT_KEY2> <DISPLAY2>
 *       ...
 *       <INPUT_KEYn> <DISPLAYn>
 *     %keyname end
 *         #
 *         # keyname section will define all the input keys and it's display format(may be 
 *         # different than the key, like in some Big5 input methods)
 *         #
 * 
 *     [%chardef begin]
 *       KEY1  PHRASE1 [PHRASE2] ... [PHRASEn]
 *       KEY2  PHRASE1 [PHRASE2] ... [PHRASEn]
 *       ...
 *       KEYm  PHRASE1 [PHRASE2] ... [PHRASEn]
 *     [%chardef end]
 *         #
 *         # This is the key->char/phrase mapping section, you can put multiple char/phrases
 *         # in the same line, separated by SPACE or TAB. Some cxterm's tit uses such format
 *         #      KEY  CHAR1CHAR2CHAR3...CHARN
 *         # There is no SPACE/TAB between chars, in such case, you need to use -charonly
 *         # command line option so cin2tab will treat it as single characters instead of
 *         # phrase.
 *         #
 *         # For compatibility, cin2tab will also support the following format:
 *         #      PhraseKey  (no space between Phrase(Chinese) and Key(ASCII)
 *         #      Phrase Key  (with SPACE or TAB between Phrase and Key)
 *         #      KeyPhrase1 [Phrase2] ...
 *         #
 *************************************************************************************************/

 o cin2tab can also (partly) support Microsoft's MaBiao format:

    [Description]
    Name=ChineseName
    MaxxCodes=n
    MaxElement=1
    UsedCodes=/abcdefghijklmnoprstuvwxyz
    WildChar=q
    NumRules=3
    [Rule]
    rule
    ...
    rule
    [Text]
    PhraseKey 
    ...
    PhraseKey

    cin2tab will ignore most of the keywords except "Name=", "UsedCodes=" and "WildChar=".
    After "[Text]", it's the actual phrase->key mapping, there is no SPACE/TAB between them.


 o  cin2tab can also support pure mapping format(without any % keyword)
    In this case, the input keys will be default ones(a-z) and select keys are 0-9.

    Format1:   Key Phrase [Phrase...]
    Format2:   KeyPhrase [Phrase...]
    Format3:   Phrase Key
    Format4:   PhraseKey